Linking Changes in Contraceptive Use to Declines in Teen Pregnancy Rates
Total Page:16
File Type:pdf, Size:1020Kb
Article Linking Changes in Contraceptive Use to Declines in Teen Pregnancy Rates Jennifer Manlove 1,*, Quentin Karpilow 1, Kate Welti 1 and Adam Thomas 1,2 Received: 5 October 2015; Accepted: 21 December 2015; Published: 24 December 2015 Academic Editor: Naomi Farber 1 Reproductive Health & Family Formation, Child Trends, Bethesda, MD 20814, USA; [email protected] (Q.K.); [email protected] (K.W.); [email protected] (A.T.) 2 McCourt School of Public Policy, Georgetown University, Washington, DC 20057, USA * Correspondence: [email protected]; Tel.: +1-240-223-9262 Abstract: Using a unique microsimulation tool, Teen FamilyScape, the present study explores how changes in the mix of contraceptive methods used by teens contributed to the decline in the U.S. teen pregnancy rate between 2002 and 2010. Results indicate that changes in contraceptive use contributed to approximately half of the decline in the teen pregnancy rate during this time period (48%) and that a little more than half of this “contraceptive effect” was due to an increase in teen condom use (58%). The remaining share of the contraceptive effect can be attributed to an increase in the use of more effective hormonal (pill, patch, ring) and long-acting reversible contraceptive (LARC)/injectable methods (Intrauterine Devices (IUD), implant and injectable). Results from an additional counterfactual analysis suggest that the contraceptive effect was driven by the fact that the percentage of teens using no birth control fell during the study time period, rather than by the fact that some teens switched from less effective methods (condoms) to more effective hormonal and LARC/injectable methods. However, very high typical use failure rates for teen condom users suggest the need for a two-pronged approach for continuing reductions in teen pregnancy for sexually active teens: first, targeting the youth most at risk of not using contraception and helping them choose contraception, and second, increasing the effectiveness of method use among existing contraceptors. Keywords: teen pregnancy; contraceptive methods; historical trends 1. Introduction The teen pregnancy rate in the U.S. has declined dramatically over the last two decades [1], declining by almost one-third (31%) between 2000 and 2010 (the most recent year for which data are available) from 83.4 pregnancies per 1000 teenaged women to 57.4 pregnancies per 1000 teenaged women. This decline is the result of two underlying trends: reductions in the percentage of teenagers who are sexually active and improvements in contraceptive use among teens who are sexually active. There has been some debate about whether recent declines in teen pregnancies are due primarily to increases in abstinence or to improvements in contraceptive use. A number of studies have attempted to parse out the relative importance of these trends. These studies generally find that both factors have contributed to the reduction in teen pregnancies. The estimated magnitudes of the abstinence and contraceptive effects, however, differ, based on the specific time period studied, the dataset used, and the way in which sexual activity and contraceptive use are measured. For example, two studies that focused primarily on trends in the 1990s found that declines in sexual activity had the greatest impact on reductions in teen pregnancy. Mohn et al.,[2] found that the decline in teens engaging in sex accounted for 67% of the drop. Meanwhile, Santelli et al.,[3] calculated that 53% of the decline in Societies 2016, 6, 1; doi:10.3390/soc6010001 www.mdpi.com/journal/societies Societies 2016, 6, 1 2 of 14 teen pregnancy rates among high school students could be attributed to a higher percentage delaying sexual initiation and the rest could be attributed to improved contraceptive use. However, more recent studies, using data into the 2000s, found that trends in contraceptive use had a greater impact on reductions in teen pregnancies or births. Santelli et al.,[4] found that improvements in contraceptive use accounted for 86% of the reduction in teen pregnancy between 1995 and 2002 for teens aged 15–19. The authors also found that improved contraceptive use accounted for 77% of the reduction in pregnancies among teens aged 15–17. Kearney and Levine [5] studied the related drop in teen births between 1991 and 2007 among high school students and found that increased contraceptive use was responsible for 65% of the decline. Additionally, a review by the Guttmacher Institute found there was no decline in the share of teens who were sexually experienced between 2003 and 2010, but found evidence of improvements in contraceptive use (reductions in non-use combined with increases in the use of more effective methods). The report concluded that the decline in the teen pregnancy rate during that time period was due primarily to improvements in contraceptive use [6]. Although estimates differ based on time period and measurement, all of these studies found that a substantial percentage of declines in teen pregnancies or births were due to changes in contraceptive use among teens. However, none of the papers described above examined which changes in contraceptive use contributed most to the overall contraceptive effect. This is an important gap in the literature because some changes in contraceptive use have greater impacts than others. For example, long-acting reversible contraceptive (LARC) methods, such as Intrauterine Devices (IUDs) and subdermal implants are notably more effective than other, more user-dependent methods such as condoms and oral contraception. Because they are less susceptible to user error, these methods’ perfect-use failure rates (the annual rates of pregnancy among women who use their methods consistently and correctly at each act of intercourse) and their typical-use failure rates (the average annual pregnancy rates accounting for the fact that many women do not always use their method correctly and consistently) are both less than 1% [7]. On the other hand, among pill users, the perfect-use failure rate is about the same as for long-acting methods, but the typical-use failure rate is higher, at 9% [7]. Among condom users, the perfect-use failure rate is only 2%, while the typical-use failure rate is 18% [7]. Studies have found that teens are less likely than adults to be perfect users of their chosen birth control method and therefore experience even higher failure rates when relying on user-dependent methods [8–10]. Therefore, increasing the percentage of teen women using highly effectively LARCs can further reduce pregnancies versus increasing the percentage that use condoms. While there is considerable variation in different methods’ failure rates, even the least effective methods can substantially reduce the risk of pregnancy, relative to the use of no method. Our analyses of the 2011–2013 wave of the National Survey of Family Growth (NSFG) suggest that 12% of teens who were sexually active in the past three months did not use contraception at last sex. The annual rate of pregnancy is estimated to be 85% among sexually active women who do not use contraception [7]. It might be possible, then, to achieve meaningful further reductions in teen pregnancy without large increases in the use of the most effective (but the least prevalent) forms of contraception. For example, previous research has found that dramatic reductions in the pregnancy rate can be achieved by increasing the use of condoms among those who are using no method of contraception [11]. We shed light on this issue by documenting the changes in teens’ use of long-acting methods and various other forms of contraception that accompanied the precipitous drop in teen pregnancies over the last decade. We then use a unique microsimulation model, Teen FamilyScape, to estimate the way in which changes in teens’ contraceptive behaviors contribute to population-wide changes in pregnancy rates. The results of these analyses allow us to assess which of these changes were the most important to the reduction in the teenage pregnancy rate between 2002 and 2010. We chose this time period in part because of recent changes in contraceptive method use Societies 2016, 6, 1 3 of 14 among teens. Additionally, our focus on these years allows us to study the more recent decline in teen pregnancy, thereby expanding on previous work that focused on the 1990s and early 2000s. 2. Methods Teen FamilyScape was developed by researchers at Child Trends, Georgetown University, and The Brookings Institution1. The model is designed to reproduce real-world fertility-related behaviors and outcomes among teenagers in the United States. We use data from a range of sources2 to ensure that we realistically simulate the rate at which teenage women have sex; the frequency with which sexually active teens use contraception; the types of male-controlled and female-controlled contraception that they use; the number of teens who switch onto and off of various contraceptive methods; the frequency with which teens using various types of contraception (or none at all) become pregnant; the share of teen pregnancies that result in live births, abortions, and fetal losses; and the gestation and postpartum infertility periods for each of these pregnancy outcomes. Figure1 diagrams Teen FamilyScape’s three simulation stages. The model has a daily periodicity, which is to say that each increment in analysis time corresponds to a single day. In the first stage of the simulation, we use the female respondent file of the 2006–2010 National Survey of Family Growth (NSFG)3—a nationally representative survey that contains extensive information on sexual activity, contraceptive use, and fertility outcomes—to populate the model with a group of teenage women whose demographic characteristics are nationally representative with respect to marital status, age, race, educational attainment, and socioeconomic status [12]. Most simulated behaviors and outcomes vary according to these demographic attributes. In the second simulation stage, we use data from the NSFG to model sexual and contraceptive behavior.